536767
research-article2014
AJMXXX10.1177/1062860614536767American Journal of Medical QualityRico et al
Article
Technology Integration Performance Assessment Using Lean Principles in Health Care
American Journal of Medical Quality 2015, Vol. 30(4) 374–381 © The Author(s) 2014 Reprints and permissions: sagepub.com/journalsPermissions.nav DOI: 10.1177/1062860614536767 ajmq.sagepub.com
Florentino Rico, MS1, Ali Yalcin, PhD1, and Edward A. Eikman, MD, FACR, FACNM, FACR2
Abstract This study assesses the impact of an automated infusion system (AIS) integration at a positron emission tomography (PET) center based on “lean thinking” principles. The authors propose a systematic measurement system that evaluates improvement in terms of the “8 wastes.” This adaptation to the health care context consisted of performance measurement before and after integration of AIS in terms of time, utilization of resources, amount of materials wasted/ saved, system variability, distances traveled, and worker strain. The authors’ observations indicate that AIS stands to be very effective in a busy PET department, such as the one in Moffitt Cancer Center, owing to its accuracy, pace, and reliability, especially after the necessary adjustments are made to reduce or eliminate the source of errors. This integration must be accompanied by a process reengineering exercise to realize the full potential of AIS in reducing waste and improving patient care and worker satisfaction. Keywords lean, process improvement, technology integration, value stream map, work analysis
This article presents system improvements resulting from integrating an automated infusion system (AIS) into the preparation and dispensing of fluorodeoxyglucose (FDG) in a positron emission tomography (PET) department using lean thinking principles. Currently, FDG doses are prepared as individual doses, stored in shielded syringes and infused manually. AIS is a mobile, shielded, multidose device that automatically measures and infuses FDG. Lean thinking, or simply lean, is widely used in the manufacturing sector to improve system efficiency, eliminate waste, and maximize value for the customer. The authors adapt lean tools to a health care delivery system to measure performance improvements and value-added activities and to establish evidence to understand how best to apply lean in health care settings through systematic measurement of an implementation’s impact on labor, materials, and service. Application of lean principles requires measurement of system improvement and integration success. Despite the proliferation and success of the applications of lean principles in the health care sector,1 the effects on employees are rarely systematically measured or are based on subjective evidence.2 The true impact of lean approaches is difficult to quantify; therefore, an accurate measurement and evaluation of the integration impact can support broader adoption and acceptance.3
This study involves 2 phases. The first phase analyzes the process flow before AIS implementation, and the second phase of the study analyzes the process with AIS in place. The authors measure and evaluate the impact of the implementation in terms of time, utilization of resources, amount of materials wasted/saved, system variability, distances traveled, and worker strain.
Background Lean can be traced back to the ideas of Henry Ford, Edward Demming, and other innovators in manufacturing system efficiency.4 Although lean principles can be applied in any organization to improve the quality, efficiency, and speed of any business process,5 their application in health care systems brings new challenges. For instance, process improvement efforts in health care systems should focus 1
Industrial and Management Systems Engineering, University of South Florida, Tampa, FL 2 H. Lee Moffitt Cancer Center and Research Institute, Tampa, FL Corresponding Author: Florentino Rico, MS, Industrial and Management Systems Engineering, University of South Florida, 4202 E Fowler Ave, ENB 118, Tampa, FL 33620-5350. Email: [email protected]
Downloaded from ajm.sagepub.com at Stockholm University Library on August 11, 2015
375
Rico et al on value to the patient and not on profit, cost savings, and efficiency at the expense of patient quality of care.6 Important contextual differences exist between the health care sector and the manufacturing sector. Radnor et al7 identified 2 critical variations of the assumptions underlying lean when applied in health service systems. First, patients and health care providers are both customers and stakeholders of the system. Second, health care is predominantly designed to be capacity driven rather than demand driven; thus, there is limited opportunity to influence demand or make full use of excess resource capacity. Implementation of lean in health care settings has resulted in improved care processes and patient outcomes. Systematic reviews by Holden,2 Glasgow et al,3 and Mazzocato et al1 provide a comprehensive overview of the applications of several quality improvement projects using lean in health care settings. The most common lean techniques used are value stream mapping (VSM), the 5S method, root cause analysis, and the A3 report, among others. These tools are used to describe the current and planned process steps, identify customer value, and standardize procedures. The most common changes resulting from lean implementations include the following: process redesign, data collection and monitoring, staffing/space reassignment, and education and training. In recent studies, lean techniques have been applied to health care systems to identify value and propose opportunities for improvement using VSM8-11; to achieve sustained gains over 4 years in turnaround time, increased testing volume, monetary savings, reduction of variability, and better space utilization12; to train health care providers with the objective of improving quality13,14; and to enhance patient and staff satisfaction while improving the quality of care delivered to patients.15
the lists of tasks and process flow maps of the PET scan process before and after AIS integration. The patient preparation segment (tasks 1-5) consists of the technologist retrieving the patient from the waiting area, measuring the patient’s weight, bringing the patient to the intravenous (IV) room, and performing the venipuncture. The FDG infusion segment (tasks 6-22), which is directly affected by AIS integration, includes the following activities:
Methods
Process Description After AIS Integration and Process Modifications
Setting The study was conducted at the Moffitt H. Lee Cancer Center’s PET department. The Cancer Center, located in Tampa, Florida, is a not-for-profit institution committed to fighting cancer by working in the areas of patient care, research, and education. PET scans are used regularly to diagnose different cancers or to check how cancer patients are responding to treatment.16 The PET department has 2 scanners, employs 5 technologists, serves patients 5 d/wk from 7 am to 5 pm, and performs an average of 25 PET scans per day.
Process Description Before AIS Integration The PET scan process is divided into 3 segments: patient preparation, FDG infusion, and scanning. Figure 1 shows
wait
for the saline solution to flow through the patient who is connected to an IV line; retrieve and measure the individual dose’s desired amount of FDG (10 mCi) in the Hot Lab; infuse FDG to the patient through the IV line; flush the syringe with saline solution from the line to get the maximum amount of FDG possible into the patient; take the syringe back to the Hot Lab to measure the amount of FDG left in the syringe (residual); and return to the IV room and wait for the patient to receive more saline solution. Finally, the scanning segment (tasks 23-31) consists of waiting for the FDG to become concentrated and bringing the patient to the scanner to perform the PET scan. Other relevant tasks affected by AIS pertain to the setup and clean-up activities that are performed during various times of the day. These tasks are as follows: Saline
solution kit setup: This includes setting up the saline solution bags with IV tubing, flushing the tubing, and hanging the bags. FDG dose receiving: The doses are received by entering each one into the information system.
After AIS implementation, approximately 1 month was allowed for the learning process regarding the use of the new technology. The new FDG infusion segment (tasks 6-19) is as follows: check
saline solution in the vein by administering 10 mL of saline solution prior to connecting to the Patient Administration Set (PAS); set up AIS to perform the FDG infusion (ie, enter patient code and prime the system to prepare it for the saline test injection); connect patient through the PAS; disconnect patient from the PAS and remove catheter; take patient to the waiting area;
Downloaded from ajm.sagepub.com at Stockholm University Library on August 11, 2015
2.27
0.34
96.82
376
Downloaded from ajm.sagepub.com at Stockholm University Library on August 11, 2015
VA T1M T2M Total
- - - - - - - - - Patient arrives - - - - - - - - - -
Output: PET Scan
Customer: Doctor
Before the AIS number of tasks Time (min) 6 39.2 20 100.6 5 9.3 31 149.1
Insurance
Send scan orders
Output: PET Scan - - - - - - - - - Patient arrives - - - - - - - - - -
5HWULHYHSDWLHQWIURPZDLWLQJDUHD %ULQJSDWLHQWWRWKHVFDQQHU +HOSSODFHSDWLHQWRQVFDQQHUEHG *RWRVFDQQHUFRPSXWHUFRQWURO ([HFXWHVFDQ :DLWIRUVFDQUHVXOWV +HOSSDWLHQWRIIVFDQQHUEHG
:DLWZKLOH )'*EHFRPHV FRQFHQWUDWHG
With the AIS number of tasks Time (min) 8 37.2 14 103.6 5 1.7 27 142.5
(5.36)
30.54
97.80
(11.33)
Insurance
Abbreviations: AIS, automated infusion system; T1M, type 1 Muda; T2M, type 2 Muda; PET, positron emission tomography.
Customer: Doctor
Invoice Send scan orders
Daily demand order
FDG Supplier: Pharmacy
Patient Master Schedule
Figure 1. Value stream mapping (VSM) and summary of impact on the VSM task classification for the process before AIS and with AIS in place.
(5.86)
),)2
5HWULHYHSDWLHQWIURPZDLWLQJDUHD %ULQJSDWLHQWWRWKHVFDQQHU +HOSSODFHSDWLHQWRQVFDQQHUEHG *RWRVFDQQHUFRPSXWHUFRQWURO ([HFXWHVFDQ :DLWIRUVFDQUHVXOWV +HOSSDWLHQWRIIVFDQQHUEHG
:DLWZKLOH )'*EHFRPHV FRQFHQWUDWHG
Invoice
1.24
(0.44)
7DNHXVHGWXELQJEDFNWRWKH+RW/DE 2SHQ,RQL]DWLRQFKDPEHUDQGSODFHWXELQJ &DOLEUDWH)'*UHVLGXDOV 5HFRUGFDOLEUDWLRQYDOXH 'LVFDUGUHVLGXDOV WXELQJ
%ULQJSDWLHQWIURP,9URRPWRZDLWLQJDUHD
(0.54)
1.16
'LVFRQQHFWSDWLHQWIURPV\VWHP 5HWULHYH)'*LQIXVLRQUHFRUG
6WDUW,QIXVLRQDQGZDLW
6WDUWGRVHH[WUDFWLRQDQGZDLW
6WDUW7HVW,QMHFWLRQDQGZDLW
(0.31)
0.48
1.03
(0.14)
0.68
(0.12)
(0.03)
),)2
32.50
(16.67)
7DNHSDWLHQWWRZDLWLQJDUHD
Patient Master Schedule
nts
(1)
tie
1.57
pa
(1.05)
h of
tem
1.19
Sc le
sys
:DLWIRU VDOLQH VROXWLRQWR IORZ
pu the
*RWR,9URRPWRUHWULHYHSDWLHQW
ily u ed nts
0.16
Re tie
(0.13)
d in into
tem
0.84
rd
co t re sys
(0.29)
an pa
7DNHXVHGV\ULQJHEDFNWRWKH+RW/DE 2SHQ,RQL]DWLRQFKDPEHUDQGSODFHV\ULQJH &DOLEUDWH)'*UHVLGXDOV 5HFRUGFDOLEUDWLRQYDOXH 'LVFDUGUHVLGXDOV 5HFRUGSDWLHQWGRVHWLPHV
Da Sc of
0.28
ily ule
(0.39)
nd 6HWXS,QWHJR3HW,QIXVLRQV\VWHP &RQQHFW3DWLHQW$GPLQLVWUDWLRQ6HWWRFDWKHWHU
Da d he
$GPLQLVWHU)'*WRSDWLHQW
Se
rs
(0.3)
&KHFN6DOLQHVROXWLRQIORZLQWKHYHLQ
3UHSDUH,9IOXLGVHW 3HUIRUPYHQLSXQFWXUH
7DNHSDWLHQWWR,9URRP
2 times/ day
R the
1.02
de pu into
(0.5)
Or
1.13
0.42
(0.22)
4.68
(2.22)
(0.15)
0.33
(0.18)
5HWULHYHSDWLHQW
0HDVXUHSDWLHQW VZHLJKW
nd
d in
5HWULHYHV\ULQJHIURP+RW/DE 3LFN)'*ER[XSDQGRSHQ 7DNHV\ULQJHRXWRIER[ 3ODFHV\ULQJHLQLRQL]DWLRQFKDPEHUIRU FDOLEUDWLRQ 5HFRUGFDOLEUDWLRQYDOXH 7DNHV\ULQJHRXWDQGSXWLQWUDQVSRUWHU 7DNHV\ULQJHWR,9URRP
Daily demand order
2.41
0.32
(0.55)
Se
an ord
1.55
e ceiv rs
:DLWIRU VDOLQH VROXWLRQWR IORZ
FDG Supplier: Pharmacy
With the AIS 3DWLHQWDUULYHV
rde
(5.06)
3UHSDUH,9IOXLGVHW 3HUIRUPYHQLSXQFWXUH
7DNHSDWLHQWWR,9URRP
0HDVXUHSDWLHQW VZHLJKW
3 times/ day
Average Time (min), (Std. Dev)
eO
5.01
(4.61)
5.68
(0.08)
0.22
(0.22)
5HWULHYHSDWLHQW
FDG Infusion Segment
(0.46)
Before the AIS
3DWLHQWDUULYHV
FDG Infusion Segment
Average Time (min), (Std. Dev)
iv ece c t re
377
Rico et al Table 1. Toyota Production System’s “8 Wastes” Adapted to the Health Care Delivery Context. Category Labor
Waste Type
Measurement
1. Wait
Duration of tasks
2. Motion
Handling of shielded boxes
3. Overprocessing 4. Transportation 5. Inventory and 6. Overproduction
Service
7. Defects
T2M tasks (VSM) Technologist travel FDG injected FDG residual Tubing and saline solution use Space utilized Use of scanners System errors
Human potential
8. Talent
Not measured systematically
Materials
Method Time and motion study analysis VSM Data collection VSM VSM Process mapping Data collection Statistical analysis Work sampling
Data collection System output analysis
Abbreviations: VSM, value stream mapping; T2M, type 2 Muda; FDG, fluorodeoxyglucose.
retrieve
the printout with the amount of FDG infused and the PAS; take the PAS to the Hot Lab and measure FDG residual; and record FDG residual amount. Before AIS, both patients were prepared for the infusion, and they were infused simultaneously. After implementation, 2 patients were prepared simultaneously but only one could be injected at a time. Integration of AIS also eliminated the need to perform saline solution kit setup and FDG dose receiving tasks, which were performed once for each patient. On the other hand, it introduced new tasks: Quality control: This includes running quality con-
trol, surveying for exposure, and replacing the FDG vial. FDG vial receiving: Two bulk vials are received by entering each one into the information system. FDG vial/saline solution replacement: The FDG vial, saline solution bag, Source Administration Set, and PAS are replaced.
Lean Measurement System Lean identifies 8 wastes of resources (Table 1) that are commonly misused in health care.17 After discussions with physicians, technologists, and the company manufacturing AIS, the authors determined the corresponding set of relevant measurements and methods needed to assess the impact of AIS. Although “waste of talent”
(waste 8) was not measured systematically, this study benefited from the input and recommendations from the physicians and technologists throughout the study. In Table 1, the authors present the 7 waste types, the measurements, and the methods to assess the system before and after the integration. VSM and work design methods were used,18 such as time and motion study, work sampling, process mapping, and statistical analysis. The VSM classification process19 is driven by 2 questions: Who is the customer? What is the value to the customer? The customers for this system are the patients who go through the FDG infusion process, the technologists who service the patients, and the physicians who interpret the scan images. To specify value from the standpoint of the end customer, interviews with technologists, patients, and physicians were carried out. For this context, value to the customer means a smooth scanning process for the patient and technologist and a clear highquality image for the physician to interpret. The authors classified the process tasks into those that are value added (VA) and non-VA. VA refers to those activities that unambiguously create value. On the other hand, any task that takes time but does not add any value is known as Muda or waste.20 Type 1 Muda (T1M) are activities that create no value but seem to be unavoidable with current technologies or production assets, and Type 2 Muda (T2M) are activities that create no value and are immediately avoidable. The concept of Muda encompasses the insidious processes, barriers, and impediments that obstruct our ability to achieve goals.6 Muda includes activities that do not add value for the patient or the health care provider.
Downloaded from ajm.sagepub.com at Stockholm University Library on August 11, 2015
378
American Journal of Medical Quality 30(4)
Results Wait: Duration of Tasks Before AIS integration, the average duration of FDG infusion segment was 11.33 minutes with a standard deviation of 5.56 minutes. After integration, the same process segment’s average duration was 6.42 minutes with a standard deviation of 1.04 minutes. This is a very significant reduction both in terms of the average duration (P < .01) of the activity (by 43%) and the observed variability (P < .01) in the activity times. Figure 2 presents a frequency histogram of the FDG infusion segment time before and with the AIS. The AIS eliminated the FDG dose receiving and saline solution kit preparation tasks, which on average take 1.3 minutes and 0.85 minutes per patient, respectively. The average number of patients serviced in the PET department was 25. Therefore, AIS integration eliminated tasks that total 53.75 minutes on an average day. On the other hand, the new tasks required by AIS technology—namely quality control, FDG Vial receiving, and FDG Vial/saline solution replacement—have durations of 16.7 minutes, 9.6 minutes, and 1.7 minutes, respectively, resulting in a total of 28 min/d. The overall time savings is 25.75 min/d.
Motion: Handling of Shielded Boxes Before AIS integration, for each patient, the technologists had to pick up the FDG dose, which is contained in a
25 With the AIS
20 Frequency
A time and motion study18 is used to collect the time duration estimates for the PET imaging procedure tasks. A time study is a continuous observation of a series of tasks with the purpose of establishing a standard time. To determine the duration of tasks, 50 cycles of the process were observed to ensure a minimum confidence level of 90%. This time study analysis was done using a stopwatch and notepad to document the observed time for each task. Work sampling18 is used to calculate the utilization of the 2 scanners in the PET department. Work sampling is a method for analyzing work by taking a large number of observations at random times over a significant period of time. The scanner utilization is defined as the total number times a scanner is randomly observed working (including setup times) versus the total number of times the scanner is observed. The number of observations was determined to ensure a minimum 90% confidence. Other measures such as the distance traveled per technologist, use of materials, and systems errors were determined by process mapping, observation, and data collection.
15
Before the AIS
10 5 0 0
5
10
-5
15
20
25
30
Time (min)
Figure 2. Fluorodeoxygluscose infusion segment time histogram before and with AIS integration. Abbreviation: AIS, automated infusion system.
shielded box weighing 10.34 kg, and transport the FDG dose a distance of 19.51 m to the IV room in another container that weights 6.35 kg. With AIS, there is no need for individual doses. Thus, technologists do not have to pick up or carry FDG containers. FDG is stored in a shielded vial inside the AIS. The vial with FDG is wheeled from the Hot Lab to the IV room. FDG doses contained in a shielded vial are placed in the AIS twice a day; this vial weighs 7.53 kg.
Overprocessing: T2M Tasks Figure 1 presents the VSM and a summary of the impact of the AIS integration on the VSM task classification with respect to the number of tasks and the duration of tasks. Before AIS integration, the PET scan process was made up of 6 VA tasks, 20 T1M tasks, and 5 T2M tasks. The VA, T1M, and T2M tasks had an estimated duration of 39.2, 100.6, and 9.3 minutes, respectively. With AIS integration, there were 8 VA tasks, 14 T1M tasks, and 5 T2M tasks. The VA, T1M, and the T2M tasks had an estimated duration of 37.2, 103.6, and 1.7 minutes, respectively. New T2M tasks associated with the integration are “check saline solution flow in the vein” and “calibrate FDG residuals left in the tubing in the hot lab.” Other T2M tasks such as travel to and from the Hot Lab to retrieve the FDG doses were eliminated.
Transportation: Technologist Travel The distance traveled by technologists within the PET department before AIS integration is shown in Table 2. Servicing each patient requires a technologist to travel a total of 132.89 m. After AIS integration, each patient requires the technologist to walk approximately 113.995 m, resulting in a 14.6% reduction.
Downloaded from ajm.sagepub.com at Stockholm University Library on August 11, 2015
379
Rico et al Table 2. Distance Traveled Per Technologist Before and With AIS. From/To
Percentage of Distance Traveled Per Patient
Count/Patient
Distance (m)
1 4 1 2
67.06 9.75 10.36 8.23 132.89
50.5% 29.4% 7.8% 12.4%
1 2 1 2
67.06 10.36 9.75 8.23 113.995
58.8% 18.2% 8.6% 14.4%
Before AIS IV room/Reception/IV room IV room/Hot lab IV room/Waiting area Scanner/Waiting area Total With AIS IV room/Reception/IV room IV room/Waiting area/IV room IV room/Hot lab Scanner/Waiting area Total
Abbreviations: AIS, automated infusion system; IV, intravenous.
the PAS further reduced the residual FDG amounts to around 0.0001 mCi on a consistent basis. Before the AIS
Inventory: Space Utilization
With the AIS
8
9
10 11 Amount of FDG mCi
12
Figure 3. Whisker-and-box plot for the amount of FDG injected before and with the AIS.
Inventory: Amount of FDG Injected The predetermined desired amount of FDG to be infused into the patient is 10 mCi; ±2 mCi is acceptable. All observed values in this study fall within this range, as shown in Figure 3. After AIS integration, technologists set up the AIS to deliver 10 mCi to the patient. Technologists record the amount of FDG injected, and this information is gathered from the printout that the AIS provides. These values range from ±0.2mCi of the desired amount of 10 mCi.
Inventory: Amount of FDG Residuals Based on 50 samples observed before AIS integration, the average FDG residual was 0.017 mCi. After AIS integration, again based on 50 observations, the average residual in the PAS was 0.0003 mCi. Midway into the study, an adjustment in the amount of saline used to flush
With AIS integration, only one vial is stored during the day. The volume of 1 FDG vial shielded box is 0.0297 m3 and takes 0.1171 m2 of floor space. The AIS is placed in the IV room where the patient is prepared for the injection. The volume that the AIS requires is 0.7079 m3 and has an area of 0.5203 m2. Although AIS requires additional space in the IV room, use of a single vial eliminates the clutter of individual doses in the Hot Lab.
Inventory: Tubing and Saline Solution Use and Cost Before the integration, a 500-mL saline solution bag and one IV tubing set were used for each patient. After the integration, two 1000-mL and one 500-mL saline solution bag and 2 Source Administration Sets are used each day. Each patient also requires 1 PAS. Before AIS integration, the total cost of saline solution and tubing per patient was $2.68; after the integration, this cost is $14.61.
Inventory: Use of PET Scanners Use of capital-intensive PET scanners is an important performance measure for the PET department. There are 2 scanners in the PET department. The scanners start working at approximately 9:30 am when the first patient is ready and keep working until the last scheduled patient is scanned. Before AIS, the average resource use was 86% for scanner 1 and 72% for scanner 2. After the
Downloaded from ajm.sagepub.com at Stockholm University Library on August 11, 2015
380
American Journal of Medical Quality 30(4)
implementation, the average resource use was 83% for scanner 1 and 73% for 2.
Overproduction Waste of overproduction is closely related to the aforementioned inventory measures. After AIS integration, overproduction is measured by the excess inventory originated: savings of saline solution and tubing, reduction of residuals, availability of space in the Hot Lab, and a more accurate amount of FGD infused into the patient.
Defects: System Errors The equipment manufacturer provided a summary of errors from the computer system’s error logs by type, date, and time. During the second phase of study, 141 errors were observed in the 25 working days after the integration. System-dependent errors made up 43.6% of the observed errors. On most days, the number of system errors was limited to less than 4.
Discussion Integration of AIS has resulted in considerable process improvements related to waste reduction in labor. These improvements are especially significant (P < .01) in reducing the FDG infusion time by approximately 50% and setup activities by approximately 26 min/d. Distance traveled was reduced 14% by eliminating the travel between the IV room and the Hot Lab, and worker strain reduced by eliminating handling of individual shielded boxes containing FDG doses. T2M activities were identified using VSM and eliminated from the process, reducing the duration of these activities by 81%. However, there are new T2M activities associated with the technologist ensuring that the AIS is working correctly (eg, remeasuring residuals in PAS and checking that the saline solution is flowing from AIS to the patient properly). Another important impact of the AIS is related to the materials used. The accuracy of the amount of FDG injected into the patients improved. The maximum deviations from the desired amount of 10 mCi were reduced from 20% to 2%. In addition, the residual FDG amounts were reduced from 0.017 mCi to almost undetectable amounts. There was no change in the cost of FDG because this is based on the number of doses delivered from the pharmacy, which is dependent on the number of patients served. Integration of the AIS reduced the amount of saline solution used but increased the total cost of tubing, resulting in a net increase in the cost of materials. Hot Lab organization improved by eliminating the clutter of individual doses, but the AIS requires additional space in the IV room. The authors did not find a significant impact
on the use of scanners with the AIS (P = .437 for scanner 1 and P = .835 for scanner 2). During the use of AIS, the technologists faced several service challenges related to the number of errors generated by the system. Although almost half of these errors were operator dependent and can be remedied by a training program, the frequency of system-dependent errors was a source of frustration for the technologists. The unexpectedly high number of system-dependent errors encountered during the study has brought to bear several improvements to the infusion system. As far as the costs considered with the scope of the study, which was limited to the PET department, the intervention increased the cost per patient. However, the ability to use a bulk FDG dose container (vs individual doses) affects the cost of FDG preparation and transportation costs from the pharmacy to the hospital. At the time of the study, these costs were not available, and therefore, the overall cost savings related to the intervention could not be calculated. Another limitation of the study is that, although this systematic measurement benefited from the continuous input and engagement of employees during AIS integration, employee engagement was not assessed systematically. Future studies and systematic measurement methods of lean intervention should evaluate employee engagement (waste 8) through the use of surveys and documentation of recommendations.
Conclusions This study establishes evidence for understanding how best to apply lean in health care settings through systematic measurement of an implementation’s impact on labor, materials, and service. The study addresses a significant gap in current lean implementations, where accurate measurement of the impact on employees and service is lacking. It focuses on quantifying improvements using systematic measurement based on lean techniques. The findings presented in this article suggest that a broader understanding of improvements in the system can be achieved through systematic measurement of the impact of AIS integration. Having tangible results facilitates the integration of lean principles into the institution’s culture of improvement and management of technological change. This lean application has no impact on the demand or the capacity of the system because health care settings are designed to be capacity driven. As with the integration of all automated systems into existing manual processes, the full potential of future technology integrations cannot be realized without making the necessary adjustments to the existing process by going through a proper process reengineering exercise to ensure that unnecessary legacy tasks are eliminated from
Downloaded from ajm.sagepub.com at Stockholm University Library on August 11, 2015
381
Rico et al the process. Often, such integrations also require reallocation of labor and equipment resources to ensure a smoothly operating facility with maximum profit. Declaration of Conflicting Interests The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported in part by University of South Florida Grant #2103102300.
References 1. Mazzocato P, Savage C, Brommels M, Aronsson H, Thor J. Lean thinking in healthcare: a realist review of the literature. Qual Saf Health Care. 2010;19:376-382. 2. Holden RJ. Lean thinking in emergency departments: a critical review. Ann Emerg Med. 2011;57:265-378. 3. Glasgow JM, Scott-Caziewell JR, Kaboli PJ. Guiding inpatient quality improvement: a systematic review of lean and six sigma. Jt Comm J Qual Patient Saf. 2010;36:533-540. 4. Womack J, Jones D. The Machine That Changed the World: The Story of Lean Production. New York, NY: Free Press; 2007:352. 5. Morris L. The Toyota way. Bus B Rev. 2004;21(12):1-10. 6. Dart RC. Can lean thinking transform American health care? Ann Emerg Med. 2011;57:279-281. 7. Radnor ZJ, Holweg M, Waring J. Lean in healthcare: the unfilled promise? Soc Sci Med. 2012;74:364-371. 8. Yousri TA, Khan Z, Chakrabarti D, Fernandes R, Wahab K. Lean thinking: can it improve the outcome of fracture neck of femur patients in a district general hospital? Injury. 2011;42:1234-1237.
9. Hoskins R. Is it time to “lean” in emergency care? Int Emerg Nurs. 2010;18:57-58. 10. King DL, Ben-Tovim DI, Bassham J. Redesigning emergency department patient flows: application of lean thinking to health care. Emerg Med Australas. 2006;18:391-397. 11. Jimmerson C, Weber D, Sobek DK. Reducing waste and errors: piloting lean principles at Intermountain Healthcare. Jt Comm J Qual Patient Saf. 2005;31:249-257. 12. Rutledge J, Xu M, Simpson J. Application of the Toyota Production System improves core laboratory operations. Am J Clin Pathol. 2010;133:24-31. 13. Morganti KG, Lovejoy S, Beckjord EB, Haviland AM, Haas AC, Farley DO. A retrospective evaluation of the perfecting patient care university training program for health care organizations. Am J Med Qual. 2014;29:30-38. 14. Weigel C, Suen W, Gupte G. Using lean methodology to teach quality improvement to internal medicine residents at a safety net hospital. Am J Med Qual. 2013;28: 392-399. 15. Dickson EW, Singh S, Cheung DS, Wyatt CC, Nugent AS. Application of lean manufacturing techniques in the emergency department. J Emerg Med. 2009;37:177-182. 16. MedlinePlus. PET scan. http://www.nlm.nih.gov/medlineplus/ency/article/003827.htm. Accessed January 10, 2014. 17. Graban M. Lean Hospitals: Improving Quality, Patient Safety, and Employee Engagement. 2nd ed. Boca Raton, FL: Taylor & Francis; 2011:31-46. 18. Freivalds A, Niebel B. Methods, Standards and Work Design. 12th ed. New York, NY: McGraw-Hill Science/ Engineering/Math; 2008:405-575. 19. Rother M, Shook J, Womack J, Jones D. Learning to See: Value Stream Mapping to Add Value and Eliminate MUDA. Cambridge, MA: Lean Enterprise Institute; 2009:5-20. Version 1. 20. Liker J. The Toyota Way: 14 Management Principles From the World’s Greatest Manufacturer. 1st ed. Ipswich, MA: McGraw-Hill Education; 2003:88-89.
Downloaded from ajm.sagepub.com at Stockholm University Library on August 11, 2015
research-article2014
AJMXXX10.1177/1062860614536767American Journal of Medical QualityRico et al
Article
Technology Integration Performance Assessment Using Lean Principles in Health Care
American Journal of Medical Quality 2015, Vol. 30(4) 374–381 © The Author(s) 2014 Reprints and permissions: sagepub.com/journalsPermissions.nav DOI: 10.1177/1062860614536767 ajmq.sagepub.com
Florentino Rico, MS1, Ali Yalcin, PhD1, and Edward A. Eikman, MD, FACR, FACNM, FACR2
Abstract This study assesses the impact of an automated infusion system (AIS) integration at a positron emission tomography (PET) center based on “lean thinking” principles. The authors propose a systematic measurement system that evaluates improvement in terms of the “8 wastes.” This adaptation to the health care context consisted of performance measurement before and after integration of AIS in terms of time, utilization of resources, amount of materials wasted/ saved, system variability, distances traveled, and worker strain. The authors’ observations indicate that AIS stands to be very effective in a busy PET department, such as the one in Moffitt Cancer Center, owing to its accuracy, pace, and reliability, especially after the necessary adjustments are made to reduce or eliminate the source of errors. This integration must be accompanied by a process reengineering exercise to realize the full potential of AIS in reducing waste and improving patient care and worker satisfaction. Keywords lean, process improvement, technology integration, value stream map, work analysis
This article presents system improvements resulting from integrating an automated infusion system (AIS) into the preparation and dispensing of fluorodeoxyglucose (FDG) in a positron emission tomography (PET) department using lean thinking principles. Currently, FDG doses are prepared as individual doses, stored in shielded syringes and infused manually. AIS is a mobile, shielded, multidose device that automatically measures and infuses FDG. Lean thinking, or simply lean, is widely used in the manufacturing sector to improve system efficiency, eliminate waste, and maximize value for the customer. The authors adapt lean tools to a health care delivery system to measure performance improvements and value-added activities and to establish evidence to understand how best to apply lean in health care settings through systematic measurement of an implementation’s impact on labor, materials, and service. Application of lean principles requires measurement of system improvement and integration success. Despite the proliferation and success of the applications of lean principles in the health care sector,1 the effects on employees are rarely systematically measured or are based on subjective evidence.2 The true impact of lean approaches is difficult to quantify; therefore, an accurate measurement and evaluation of the integration impact can support broader adoption and acceptance.3
This study involves 2 phases. The first phase analyzes the process flow before AIS implementation, and the second phase of the study analyzes the process with AIS in place. The authors measure and evaluate the impact of the implementation in terms of time, utilization of resources, amount of materials wasted/saved, system variability, distances traveled, and worker strain.
Background Lean can be traced back to the ideas of Henry Ford, Edward Demming, and other innovators in manufacturing system efficiency.4 Although lean principles can be applied in any organization to improve the quality, efficiency, and speed of any business process,5 their application in health care systems brings new challenges. For instance, process improvement efforts in health care systems should focus 1
Industrial and Management Systems Engineering, University of South Florida, Tampa, FL 2 H. Lee Moffitt Cancer Center and Research Institute, Tampa, FL Corresponding Author: Florentino Rico, MS, Industrial and Management Systems Engineering, University of South Florida, 4202 E Fowler Ave, ENB 118, Tampa, FL 33620-5350. Email: [email protected]
Downloaded from ajm.sagepub.com at Stockholm University Library on August 11, 2015
375
Rico et al on value to the patient and not on profit, cost savings, and efficiency at the expense of patient quality of care.6 Important contextual differences exist between the health care sector and the manufacturing sector. Radnor et al7 identified 2 critical variations of the assumptions underlying lean when applied in health service systems. First, patients and health care providers are both customers and stakeholders of the system. Second, health care is predominantly designed to be capacity driven rather than demand driven; thus, there is limited opportunity to influence demand or make full use of excess resource capacity. Implementation of lean in health care settings has resulted in improved care processes and patient outcomes. Systematic reviews by Holden,2 Glasgow et al,3 and Mazzocato et al1 provide a comprehensive overview of the applications of several quality improvement projects using lean in health care settings. The most common lean techniques used are value stream mapping (VSM), the 5S method, root cause analysis, and the A3 report, among others. These tools are used to describe the current and planned process steps, identify customer value, and standardize procedures. The most common changes resulting from lean implementations include the following: process redesign, data collection and monitoring, staffing/space reassignment, and education and training. In recent studies, lean techniques have been applied to health care systems to identify value and propose opportunities for improvement using VSM8-11; to achieve sustained gains over 4 years in turnaround time, increased testing volume, monetary savings, reduction of variability, and better space utilization12; to train health care providers with the objective of improving quality13,14; and to enhance patient and staff satisfaction while improving the quality of care delivered to patients.15
the lists of tasks and process flow maps of the PET scan process before and after AIS integration. The patient preparation segment (tasks 1-5) consists of the technologist retrieving the patient from the waiting area, measuring the patient’s weight, bringing the patient to the intravenous (IV) room, and performing the venipuncture. The FDG infusion segment (tasks 6-22), which is directly affected by AIS integration, includes the following activities:
Methods
Process Description After AIS Integration and Process Modifications
Setting The study was conducted at the Moffitt H. Lee Cancer Center’s PET department. The Cancer Center, located in Tampa, Florida, is a not-for-profit institution committed to fighting cancer by working in the areas of patient care, research, and education. PET scans are used regularly to diagnose different cancers or to check how cancer patients are responding to treatment.16 The PET department has 2 scanners, employs 5 technologists, serves patients 5 d/wk from 7 am to 5 pm, and performs an average of 25 PET scans per day.
Process Description Before AIS Integration The PET scan process is divided into 3 segments: patient preparation, FDG infusion, and scanning. Figure 1 shows
wait
for the saline solution to flow through the patient who is connected to an IV line; retrieve and measure the individual dose’s desired amount of FDG (10 mCi) in the Hot Lab; infuse FDG to the patient through the IV line; flush the syringe with saline solution from the line to get the maximum amount of FDG possible into the patient; take the syringe back to the Hot Lab to measure the amount of FDG left in the syringe (residual); and return to the IV room and wait for the patient to receive more saline solution. Finally, the scanning segment (tasks 23-31) consists of waiting for the FDG to become concentrated and bringing the patient to the scanner to perform the PET scan. Other relevant tasks affected by AIS pertain to the setup and clean-up activities that are performed during various times of the day. These tasks are as follows: Saline
solution kit setup: This includes setting up the saline solution bags with IV tubing, flushing the tubing, and hanging the bags. FDG dose receiving: The doses are received by entering each one into the information system.
After AIS implementation, approximately 1 month was allowed for the learning process regarding the use of the new technology. The new FDG infusion segment (tasks 6-19) is as follows: check
saline solution in the vein by administering 10 mL of saline solution prior to connecting to the Patient Administration Set (PAS); set up AIS to perform the FDG infusion (ie, enter patient code and prime the system to prepare it for the saline test injection); connect patient through the PAS; disconnect patient from the PAS and remove catheter; take patient to the waiting area;
Downloaded from ajm.sagepub.com at Stockholm University Library on August 11, 2015
2.27
0.34
96.82
376
Downloaded from ajm.sagepub.com at Stockholm University Library on August 11, 2015
VA T1M T2M Total
- - - - - - - - - Patient arrives - - - - - - - - - -
Output: PET Scan
Customer: Doctor
Before the AIS number of tasks Time (min) 6 39.2 20 100.6 5 9.3 31 149.1
Insurance
Send scan orders
Output: PET Scan - - - - - - - - - Patient arrives - - - - - - - - - -
5HWULHYHSDWLHQWIURPZDLWLQJDUHD %ULQJSDWLHQWWRWKHVFDQQHU +HOSSODFHSDWLHQWRQVFDQQHUEHG *RWRVFDQQHUFRPSXWHUFRQWURO ([HFXWHVFDQ :DLWIRUVFDQUHVXOWV +HOSSDWLHQWRIIVFDQQHUEHG
:DLWZKLOH )'*EHFRPHV FRQFHQWUDWHG
With the AIS number of tasks Time (min) 8 37.2 14 103.6 5 1.7 27 142.5
(5.36)
30.54
97.80
(11.33)
Insurance
Abbreviations: AIS, automated infusion system; T1M, type 1 Muda; T2M, type 2 Muda; PET, positron emission tomography.
Customer: Doctor
Invoice Send scan orders
Daily demand order
FDG Supplier: Pharmacy
Patient Master Schedule
Figure 1. Value stream mapping (VSM) and summary of impact on the VSM task classification for the process before AIS and with AIS in place.
(5.86)
),)2
5HWULHYHSDWLHQWIURPZDLWLQJDUHD %ULQJSDWLHQWWRWKHVFDQQHU +HOSSODFHSDWLHQWRQVFDQQHUEHG *RWRVFDQQHUFRPSXWHUFRQWURO ([HFXWHVFDQ :DLWIRUVFDQUHVXOWV +HOSSDWLHQWRIIVFDQQHUEHG
:DLWZKLOH )'*EHFRPHV FRQFHQWUDWHG
Invoice
1.24
(0.44)
7DNHXVHGWXELQJEDFNWRWKH+RW/DE 2SHQ,RQL]DWLRQFKDPEHUDQGSODFHWXELQJ &DOLEUDWH)'*UHVLGXDOV 5HFRUGFDOLEUDWLRQYDOXH 'LVFDUGUHVLGXDOV WXELQJ
%ULQJSDWLHQWIURP,9URRPWRZDLWLQJDUHD
(0.54)
1.16
'LVFRQQHFWSDWLHQWIURPV\VWHP 5HWULHYH)'*LQIXVLRQUHFRUG
6WDUW,QIXVLRQDQGZDLW
6WDUWGRVHH[WUDFWLRQDQGZDLW
6WDUW7HVW,QMHFWLRQDQGZDLW
(0.31)
0.48
1.03
(0.14)
0.68
(0.12)
(0.03)
),)2
32.50
(16.67)
7DNHSDWLHQWWRZDLWLQJDUHD
Patient Master Schedule
nts
(1)
tie
1.57
pa
(1.05)
h of
tem
1.19
Sc le
sys
:DLWIRU VDOLQH VROXWLRQWR IORZ
pu the
*RWR,9URRPWRUHWULHYHSDWLHQW
ily u ed nts
0.16
Re tie
(0.13)
d in into
tem
0.84
rd
co t re sys
(0.29)
an pa
7DNHXVHGV\ULQJHEDFNWRWKH+RW/DE 2SHQ,RQL]DWLRQFKDPEHUDQGSODFHV\ULQJH &DOLEUDWH)'*UHVLGXDOV 5HFRUGFDOLEUDWLRQYDOXH 'LVFDUGUHVLGXDOV 5HFRUGSDWLHQWGRVHWLPHV
Da Sc of
0.28
ily ule
(0.39)
nd 6HWXS,QWHJR3HW,QIXVLRQV\VWHP &RQQHFW3DWLHQW$GPLQLVWUDWLRQ6HWWRFDWKHWHU
Da d he
$GPLQLVWHU)'*WRSDWLHQW
Se
rs
(0.3)
&KHFN6DOLQHVROXWLRQIORZLQWKHYHLQ
3UHSDUH,9IOXLGVHW 3HUIRUPYHQLSXQFWXUH
7DNHSDWLHQWWR,9URRP
2 times/ day
R the
1.02
de pu into
(0.5)
Or
1.13
0.42
(0.22)
4.68
(2.22)
(0.15)
0.33
(0.18)
5HWULHYHSDWLHQW
0HDVXUHSDWLHQW VZHLJKW
nd
d in
5HWULHYHV\ULQJHIURP+RW/DE 3LFN)'*ER[XSDQGRSHQ 7DNHV\ULQJHRXWRIER[ 3ODFHV\ULQJHLQLRQL]DWLRQFKDPEHUIRU FDOLEUDWLRQ 5HFRUGFDOLEUDWLRQYDOXH 7DNHV\ULQJHRXWDQGSXWLQWUDQVSRUWHU 7DNHV\ULQJHWR,9URRP
Daily demand order
2.41
0.32
(0.55)
Se
an ord
1.55
e ceiv rs
:DLWIRU VDOLQH VROXWLRQWR IORZ
FDG Supplier: Pharmacy
With the AIS 3DWLHQWDUULYHV
rde
(5.06)
3UHSDUH,9IOXLGVHW 3HUIRUPYHQLSXQFWXUH
7DNHSDWLHQWWR,9URRP
0HDVXUHSDWLHQW VZHLJKW
3 times/ day
Average Time (min), (Std. Dev)
eO
5.01
(4.61)
5.68
(0.08)
0.22
(0.22)
5HWULHYHSDWLHQW
FDG Infusion Segment
(0.46)
Before the AIS
3DWLHQWDUULYHV
FDG Infusion Segment
Average Time (min), (Std. Dev)
iv ece c t re
377
Rico et al Table 1. Toyota Production System’s “8 Wastes” Adapted to the Health Care Delivery Context. Category Labor
Waste Type
Measurement
1. Wait
Duration of tasks
2. Motion
Handling of shielded boxes
3. Overprocessing 4. Transportation 5. Inventory and 6. Overproduction
Service
7. Defects
T2M tasks (VSM) Technologist travel FDG injected FDG residual Tubing and saline solution use Space utilized Use of scanners System errors
Human potential
8. Talent
Not measured systematically
Materials
Method Time and motion study analysis VSM Data collection VSM VSM Process mapping Data collection Statistical analysis Work sampling
Data collection System output analysis
Abbreviations: VSM, value stream mapping; T2M, type 2 Muda; FDG, fluorodeoxyglucose.
retrieve
the printout with the amount of FDG infused and the PAS; take the PAS to the Hot Lab and measure FDG residual; and record FDG residual amount. Before AIS, both patients were prepared for the infusion, and they were infused simultaneously. After implementation, 2 patients were prepared simultaneously but only one could be injected at a time. Integration of AIS also eliminated the need to perform saline solution kit setup and FDG dose receiving tasks, which were performed once for each patient. On the other hand, it introduced new tasks: Quality control: This includes running quality con-
trol, surveying for exposure, and replacing the FDG vial. FDG vial receiving: Two bulk vials are received by entering each one into the information system. FDG vial/saline solution replacement: The FDG vial, saline solution bag, Source Administration Set, and PAS are replaced.
Lean Measurement System Lean identifies 8 wastes of resources (Table 1) that are commonly misused in health care.17 After discussions with physicians, technologists, and the company manufacturing AIS, the authors determined the corresponding set of relevant measurements and methods needed to assess the impact of AIS. Although “waste of talent”
(waste 8) was not measured systematically, this study benefited from the input and recommendations from the physicians and technologists throughout the study. In Table 1, the authors present the 7 waste types, the measurements, and the methods to assess the system before and after the integration. VSM and work design methods were used,18 such as time and motion study, work sampling, process mapping, and statistical analysis. The VSM classification process19 is driven by 2 questions: Who is the customer? What is the value to the customer? The customers for this system are the patients who go through the FDG infusion process, the technologists who service the patients, and the physicians who interpret the scan images. To specify value from the standpoint of the end customer, interviews with technologists, patients, and physicians were carried out. For this context, value to the customer means a smooth scanning process for the patient and technologist and a clear highquality image for the physician to interpret. The authors classified the process tasks into those that are value added (VA) and non-VA. VA refers to those activities that unambiguously create value. On the other hand, any task that takes time but does not add any value is known as Muda or waste.20 Type 1 Muda (T1M) are activities that create no value but seem to be unavoidable with current technologies or production assets, and Type 2 Muda (T2M) are activities that create no value and are immediately avoidable. The concept of Muda encompasses the insidious processes, barriers, and impediments that obstruct our ability to achieve goals.6 Muda includes activities that do not add value for the patient or the health care provider.
Downloaded from ajm.sagepub.com at Stockholm University Library on August 11, 2015
378
American Journal of Medical Quality 30(4)
Results Wait: Duration of Tasks Before AIS integration, the average duration of FDG infusion segment was 11.33 minutes with a standard deviation of 5.56 minutes. After integration, the same process segment’s average duration was 6.42 minutes with a standard deviation of 1.04 minutes. This is a very significant reduction both in terms of the average duration (P < .01) of the activity (by 43%) and the observed variability (P < .01) in the activity times. Figure 2 presents a frequency histogram of the FDG infusion segment time before and with the AIS. The AIS eliminated the FDG dose receiving and saline solution kit preparation tasks, which on average take 1.3 minutes and 0.85 minutes per patient, respectively. The average number of patients serviced in the PET department was 25. Therefore, AIS integration eliminated tasks that total 53.75 minutes on an average day. On the other hand, the new tasks required by AIS technology—namely quality control, FDG Vial receiving, and FDG Vial/saline solution replacement—have durations of 16.7 minutes, 9.6 minutes, and 1.7 minutes, respectively, resulting in a total of 28 min/d. The overall time savings is 25.75 min/d.
Motion: Handling of Shielded Boxes Before AIS integration, for each patient, the technologists had to pick up the FDG dose, which is contained in a
25 With the AIS
20 Frequency
A time and motion study18 is used to collect the time duration estimates for the PET imaging procedure tasks. A time study is a continuous observation of a series of tasks with the purpose of establishing a standard time. To determine the duration of tasks, 50 cycles of the process were observed to ensure a minimum confidence level of 90%. This time study analysis was done using a stopwatch and notepad to document the observed time for each task. Work sampling18 is used to calculate the utilization of the 2 scanners in the PET department. Work sampling is a method for analyzing work by taking a large number of observations at random times over a significant period of time. The scanner utilization is defined as the total number times a scanner is randomly observed working (including setup times) versus the total number of times the scanner is observed. The number of observations was determined to ensure a minimum 90% confidence. Other measures such as the distance traveled per technologist, use of materials, and systems errors were determined by process mapping, observation, and data collection.
15
Before the AIS
10 5 0 0
5
10
-5
15
20
25
30
Time (min)
Figure 2. Fluorodeoxygluscose infusion segment time histogram before and with AIS integration. Abbreviation: AIS, automated infusion system.
shielded box weighing 10.34 kg, and transport the FDG dose a distance of 19.51 m to the IV room in another container that weights 6.35 kg. With AIS, there is no need for individual doses. Thus, technologists do not have to pick up or carry FDG containers. FDG is stored in a shielded vial inside the AIS. The vial with FDG is wheeled from the Hot Lab to the IV room. FDG doses contained in a shielded vial are placed in the AIS twice a day; this vial weighs 7.53 kg.
Overprocessing: T2M Tasks Figure 1 presents the VSM and a summary of the impact of the AIS integration on the VSM task classification with respect to the number of tasks and the duration of tasks. Before AIS integration, the PET scan process was made up of 6 VA tasks, 20 T1M tasks, and 5 T2M tasks. The VA, T1M, and T2M tasks had an estimated duration of 39.2, 100.6, and 9.3 minutes, respectively. With AIS integration, there were 8 VA tasks, 14 T1M tasks, and 5 T2M tasks. The VA, T1M, and the T2M tasks had an estimated duration of 37.2, 103.6, and 1.7 minutes, respectively. New T2M tasks associated with the integration are “check saline solution flow in the vein” and “calibrate FDG residuals left in the tubing in the hot lab.” Other T2M tasks such as travel to and from the Hot Lab to retrieve the FDG doses were eliminated.
Transportation: Technologist Travel The distance traveled by technologists within the PET department before AIS integration is shown in Table 2. Servicing each patient requires a technologist to travel a total of 132.89 m. After AIS integration, each patient requires the technologist to walk approximately 113.995 m, resulting in a 14.6% reduction.
Downloaded from ajm.sagepub.com at Stockholm University Library on August 11, 2015
379
Rico et al Table 2. Distance Traveled Per Technologist Before and With AIS. From/To
Percentage of Distance Traveled Per Patient
Count/Patient
Distance (m)
1 4 1 2
67.06 9.75 10.36 8.23 132.89
50.5% 29.4% 7.8% 12.4%
1 2 1 2
67.06 10.36 9.75 8.23 113.995
58.8% 18.2% 8.6% 14.4%
Before AIS IV room/Reception/IV room IV room/Hot lab IV room/Waiting area Scanner/Waiting area Total With AIS IV room/Reception/IV room IV room/Waiting area/IV room IV room/Hot lab Scanner/Waiting area Total
Abbreviations: AIS, automated infusion system; IV, intravenous.
the PAS further reduced the residual FDG amounts to around 0.0001 mCi on a consistent basis. Before the AIS
Inventory: Space Utilization
With the AIS
8
9
10 11 Amount of FDG mCi
12
Figure 3. Whisker-and-box plot for the amount of FDG injected before and with the AIS.
Inventory: Amount of FDG Injected The predetermined desired amount of FDG to be infused into the patient is 10 mCi; ±2 mCi is acceptable. All observed values in this study fall within this range, as shown in Figure 3. After AIS integration, technologists set up the AIS to deliver 10 mCi to the patient. Technologists record the amount of FDG injected, and this information is gathered from the printout that the AIS provides. These values range from ±0.2mCi of the desired amount of 10 mCi.
Inventory: Amount of FDG Residuals Based on 50 samples observed before AIS integration, the average FDG residual was 0.017 mCi. After AIS integration, again based on 50 observations, the average residual in the PAS was 0.0003 mCi. Midway into the study, an adjustment in the amount of saline used to flush
With AIS integration, only one vial is stored during the day. The volume of 1 FDG vial shielded box is 0.0297 m3 and takes 0.1171 m2 of floor space. The AIS is placed in the IV room where the patient is prepared for the injection. The volume that the AIS requires is 0.7079 m3 and has an area of 0.5203 m2. Although AIS requires additional space in the IV room, use of a single vial eliminates the clutter of individual doses in the Hot Lab.
Inventory: Tubing and Saline Solution Use and Cost Before the integration, a 500-mL saline solution bag and one IV tubing set were used for each patient. After the integration, two 1000-mL and one 500-mL saline solution bag and 2 Source Administration Sets are used each day. Each patient also requires 1 PAS. Before AIS integration, the total cost of saline solution and tubing per patient was $2.68; after the integration, this cost is $14.61.
Inventory: Use of PET Scanners Use of capital-intensive PET scanners is an important performance measure for the PET department. There are 2 scanners in the PET department. The scanners start working at approximately 9:30 am when the first patient is ready and keep working until the last scheduled patient is scanned. Before AIS, the average resource use was 86% for scanner 1 and 72% for scanner 2. After the
Downloaded from ajm.sagepub.com at Stockholm University Library on August 11, 2015
380
American Journal of Medical Quality 30(4)
implementation, the average resource use was 83% for scanner 1 and 73% for 2.
Overproduction Waste of overproduction is closely related to the aforementioned inventory measures. After AIS integration, overproduction is measured by the excess inventory originated: savings of saline solution and tubing, reduction of residuals, availability of space in the Hot Lab, and a more accurate amount of FGD infused into the patient.
Defects: System Errors The equipment manufacturer provided a summary of errors from the computer system’s error logs by type, date, and time. During the second phase of study, 141 errors were observed in the 25 working days after the integration. System-dependent errors made up 43.6% of the observed errors. On most days, the number of system errors was limited to less than 4.
Discussion Integration of AIS has resulted in considerable process improvements related to waste reduction in labor. These improvements are especially significant (P < .01) in reducing the FDG infusion time by approximately 50% and setup activities by approximately 26 min/d. Distance traveled was reduced 14% by eliminating the travel between the IV room and the Hot Lab, and worker strain reduced by eliminating handling of individual shielded boxes containing FDG doses. T2M activities were identified using VSM and eliminated from the process, reducing the duration of these activities by 81%. However, there are new T2M activities associated with the technologist ensuring that the AIS is working correctly (eg, remeasuring residuals in PAS and checking that the saline solution is flowing from AIS to the patient properly). Another important impact of the AIS is related to the materials used. The accuracy of the amount of FDG injected into the patients improved. The maximum deviations from the desired amount of 10 mCi were reduced from 20% to 2%. In addition, the residual FDG amounts were reduced from 0.017 mCi to almost undetectable amounts. There was no change in the cost of FDG because this is based on the number of doses delivered from the pharmacy, which is dependent on the number of patients served. Integration of the AIS reduced the amount of saline solution used but increased the total cost of tubing, resulting in a net increase in the cost of materials. Hot Lab organization improved by eliminating the clutter of individual doses, but the AIS requires additional space in the IV room. The authors did not find a significant impact
on the use of scanners with the AIS (P = .437 for scanner 1 and P = .835 for scanner 2). During the use of AIS, the technologists faced several service challenges related to the number of errors generated by the system. Although almost half of these errors were operator dependent and can be remedied by a training program, the frequency of system-dependent errors was a source of frustration for the technologists. The unexpectedly high number of system-dependent errors encountered during the study has brought to bear several improvements to the infusion system. As far as the costs considered with the scope of the study, which was limited to the PET department, the intervention increased the cost per patient. However, the ability to use a bulk FDG dose container (vs individual doses) affects the cost of FDG preparation and transportation costs from the pharmacy to the hospital. At the time of the study, these costs were not available, and therefore, the overall cost savings related to the intervention could not be calculated. Another limitation of the study is that, although this systematic measurement benefited from the continuous input and engagement of employees during AIS integration, employee engagement was not assessed systematically. Future studies and systematic measurement methods of lean intervention should evaluate employee engagement (waste 8) through the use of surveys and documentation of recommendations.
Conclusions This study establishes evidence for understanding how best to apply lean in health care settings through systematic measurement of an implementation’s impact on labor, materials, and service. The study addresses a significant gap in current lean implementations, where accurate measurement of the impact on employees and service is lacking. It focuses on quantifying improvements using systematic measurement based on lean techniques. The findings presented in this article suggest that a broader understanding of improvements in the system can be achieved through systematic measurement of the impact of AIS integration. Having tangible results facilitates the integration of lean principles into the institution’s culture of improvement and management of technological change. This lean application has no impact on the demand or the capacity of the system because health care settings are designed to be capacity driven. As with the integration of all automated systems into existing manual processes, the full potential of future technology integrations cannot be realized without making the necessary adjustments to the existing process by going through a proper process reengineering exercise to ensure that unnecessary legacy tasks are eliminated from
Downloaded from ajm.sagepub.com at Stockholm University Library on August 11, 2015
381
Rico et al the process. Often, such integrations also require reallocation of labor and equipment resources to ensure a smoothly operating facility with maximum profit. Declaration of Conflicting Interests The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported in part by University of South Florida Grant #2103102300.
References 1. Mazzocato P, Savage C, Brommels M, Aronsson H, Thor J. Lean thinking in healthcare: a realist review of the literature. Qual Saf Health Care. 2010;19:376-382. 2. Holden RJ. Lean thinking in emergency departments: a critical review. Ann Emerg Med. 2011;57:265-378. 3. Glasgow JM, Scott-Caziewell JR, Kaboli PJ. Guiding inpatient quality improvement: a systematic review of lean and six sigma. Jt Comm J Qual Patient Saf. 2010;36:533-540. 4. Womack J, Jones D. The Machine That Changed the World: The Story of Lean Production. New York, NY: Free Press; 2007:352. 5. Morris L. The Toyota way. Bus B Rev. 2004;21(12):1-10. 6. Dart RC. Can lean thinking transform American health care? Ann Emerg Med. 2011;57:279-281. 7. Radnor ZJ, Holweg M, Waring J. Lean in healthcare: the unfilled promise? Soc Sci Med. 2012;74:364-371. 8. Yousri TA, Khan Z, Chakrabarti D, Fernandes R, Wahab K. Lean thinking: can it improve the outcome of fracture neck of femur patients in a district general hospital? Injury. 2011;42:1234-1237.
9. Hoskins R. Is it time to “lean” in emergency care? Int Emerg Nurs. 2010;18:57-58. 10. King DL, Ben-Tovim DI, Bassham J. Redesigning emergency department patient flows: application of lean thinking to health care. Emerg Med Australas. 2006;18:391-397. 11. Jimmerson C, Weber D, Sobek DK. Reducing waste and errors: piloting lean principles at Intermountain Healthcare. Jt Comm J Qual Patient Saf. 2005;31:249-257. 12. Rutledge J, Xu M, Simpson J. Application of the Toyota Production System improves core laboratory operations. Am J Clin Pathol. 2010;133:24-31. 13. Morganti KG, Lovejoy S, Beckjord EB, Haviland AM, Haas AC, Farley DO. A retrospective evaluation of the perfecting patient care university training program for health care organizations. Am J Med Qual. 2014;29:30-38. 14. Weigel C, Suen W, Gupte G. Using lean methodology to teach quality improvement to internal medicine residents at a safety net hospital. Am J Med Qual. 2013;28: 392-399. 15. Dickson EW, Singh S, Cheung DS, Wyatt CC, Nugent AS. Application of lean manufacturing techniques in the emergency department. J Emerg Med. 2009;37:177-182. 16. MedlinePlus. PET scan. http://www.nlm.nih.gov/medlineplus/ency/article/003827.htm. Accessed January 10, 2014. 17. Graban M. Lean Hospitals: Improving Quality, Patient Safety, and Employee Engagement. 2nd ed. Boca Raton, FL: Taylor & Francis; 2011:31-46. 18. Freivalds A, Niebel B. Methods, Standards and Work Design. 12th ed. New York, NY: McGraw-Hill Science/ Engineering/Math; 2008:405-575. 19. Rother M, Shook J, Womack J, Jones D. Learning to See: Value Stream Mapping to Add Value and Eliminate MUDA. Cambridge, MA: Lean Enterprise Institute; 2009:5-20. Version 1. 20. Liker J. The Toyota Way: 14 Management Principles From the World’s Greatest Manufacturer. 1st ed. Ipswich, MA: McGraw-Hill Education; 2003:88-89.
Downloaded from ajm.sagepub.com at Stockholm University Library on August 11, 2015
